基于模糊自组织神经网络的多目标跟踪算法

An Algorithm for Multiple Targets Tracking Based on Fuzzy Self-Organizing Neural Network

分析了多目标跟踪问题的数据关联算法复杂性,研究了基于模糊C均值聚类算法的模糊自组织神经网络的特性及其在多目标跟踪中的应用,提出了将FKCN算法、自组织神经网络与数据关联、滤波相结合的跟踪算法.仿真结果表明本算法能在多目标环境下取得较好的跟踪效果.

The complication feature of data association in multiple targets tracking is analyzed. The features of the fuzzy self-organizing neural network based on the fuzzy C-mean algorithm and its application in the data association for multiple targets tracking are studied, and a new tracking algorithm is proposed, which integrates the classification function of FKCN algorithm and the self-organizing neural network in data association and kalman filter estimation method. Simulation results show that the tracking algorithm works well in multiple targets tracking condition.